The VP2 problem is interesting. I didn't touch the Shader code in this release. I'll have a closer look.
Yes, a coincidence - windows updated my drivers the same night I switched to .35. However I reverted back to the Autodesk certified drivers for a Quadro 2000 and Maya 2015 (332.21) and Corona .34 and the problem is there. I thought it might be an error on my end, but CoronaSurface is the only shader doing this. VRayMtl, lambert, mia all behave under VP2. I also notice that CoronaSurface is significantly slower (at least on a Quadro) than other shaders but again it may just be me.
From Ondra, I got the information that a simple exposure of -5 results in a correct illumination, it is supposed to be overbright with default exposure.
Yes, my error - I accidentally had Simple exposure AND Sky multiplier set to 1 which produces very overexposed images. Maybe the sky multiplier should be default 1 because it really confused me that the default is actually .1?
Would it be possible to give me an example of the HDRI/physical sky problem and "simple expsure"?
I have attached two renders. One is using "Simple Exposure" other "camera" defined exposure. Both have exact same sun + sky settings and I matched both exposures via the sky as close as possible. Yet in both renders there is a BIG discrepancy in reflection strength and the GI under the eaves of the building. I don't expect to be such a difference in rendering (unless I'm interpreting Corona or some more settings wrong). :)
I have created a test scene, when rendering notice the difference in the fresnel reflections on the glass sphere and also the shadow strength from the cube (when toggling simple exposure on and off).
(I render to color gamma of 1 and use Maya's new color management feature in Maya's render view which seems to work really well with Corona if you render the attached scene)